期刊论文

Meng, T

英文

Knowledge-Based Systems

0950-7051

2025

325

113974

10.1016/j.knosys.2025.113974

CRediT authorship contribution statement Wei Ai: Writing – original draft, Validation, Supervision, Methodology, Investigation, Conceptualization. Yulu Liu: Writing – original draft, Visualization, Validation, Software, Methodology, Investigation, Conceptualization. Chen Wei: Writing – review & editing, Validation, Supervision, Project administration. Tao Meng: Writing – review & editing, Supervision, Project administration, acquisition. Hongen Shao: Writing – review & editing, Visualization, Software, Resources, Methodology,

本校为第一且通讯机构

Entity alignment (EA) is the task of identifying equivalent entities in two knowledge graphs (KGs) using a limited set of seed entities. Existing research mainly uses graph neural networks (GNNs) to aggregate entity neighborhood features for representation to achieve better entity alignment. However, most of them ignore the fusion of multiple relations between entities and the mining of latent relations, which limits the effectiveness of entity representation to some extent. Therefore, this paper proposes a novel multi-relation fusion and latent-relation mining graph convolutional network (MFL...